1. Field of the Invention
The present invention is generally related to semiconductor processing. In particular, the present invention relates to processing ferroelectric films.
2. Description of the Related Art
A ferroelectric material is a material that exhibits an ability to maintain an electric polarization in the absence of an applied electric field. Ferroelectric materials also exhibit piezoelectricity, where the material changes polarization in response to a mechanical pressure or strain, and pyroelectricity, where the material changes polarization in response to a temperature change.
The foregoing properties of ferroelectric materials have led to many practical applications. One application uses the ability of the ferroelectric material to retain a polarization state to store data in a non-volatile memory device.
A film of zinc oxide (ZnO) doped with lithium (Li) and/or magnesium (Mg) is known to form a weak ferroelectric film, Akira Onodera, et al., Ferroelectric Properties in Piezoelectric Semiconductor Zn1xe2x88x92xMxO (M=Li, Mg), 36 Japan J. Appl. Phys. 6008 (1997). In a prior patent application, Applicants disclosed a non-volatile semiconductor memory fabricated from a doped ZnO film, Weak Ferroelectric Memory Transistor, Application No. 09/383,726, filed Aug. 26, 1999, the entirety of which is hereby incorporated by reference.
ZnO in stoichiometric form is an electrical insulator. Conventional methods of doping host ZnO with Li and/or Mg to form ferroelectric films have proven inadequate. Conventional methods are not well suited to the doping of host ZnO with Li and/or Mg for relatively large scale operations with wafers of approximately 300 millimeters (about 12 inches) or larger.
Magnetron sputtering is a conventional method of doping ZnO with Li and/or Mg. In magnetron sputtering, a target produced from a composition of ZnO and Li and/or Mg is introduced into a sputtering system. The composition can be made from ZnO with strips or particles of Li and/or Mg. Powder metallurgy can also be used to create the target.
A magnetron sputtering system creates a plasma, which reacts with the surface of the target to create the film. Disadvantageously, the film composition cannot be finetuned because the doping levels of Li and/or Mg are dictated by the initial composition of the target. Another disadvantage of magnetron sputtering is that relatively large targets, such as 300-millimeter targets, are relatively difficult to produce using powder metallurgy. A further disadvantage of magnetron sputtering with powder metallurgy is that the purity of ZnO in a powdered metal target process is relatively lower than the purity of ZnO that is attainable from a zone-refined process.
Jet vapor deposition (JVD) is another conventional method of forming a ZnO film (with or without doping of Li and/or Mg) on a substrate. In a JVD process, jets of a light carrier gas, such as helium, transport the depositing vapor of ZnO to the substrate. Uniformity of the thickness of the deposited film can require the JVD process to move and rotate the substrate relative to the jet nozzles in a complex mechanical motion. The chamber performing the JVD process can quickly grow to relatively large and expensive proportions as the chamber holding the substrate should be at least twice the diameter of the substrate wafer to accommodate the complex mechanical motion. Further, a JVD process does not permit the tailoring of the Li and/or Mg doping of the ZnO to conform the ZnO film to a desired ferroelectric characteristic.
Low-pressure chemical vapor deposition (LP-CVD) is still another conventional method of growing a ZnO film, with or without Li and/or Mg doping, on a substrate. As with the above-noted processes, the LP-CVD process also does not permit the tailoring of the Li and/or Mg doping of the ZnO to conform the ZnO film to a desired ferroelectric characteristic is not easy.
Thus, conventional methods are not well adapted to produce ferroelectric films on large wafers. The processing of ferroelectric films on large wafers of substrate can dramatically reduce a per-unit cost of chips made from the wafers. Conventional methods are also not well adapted to permit the uniform tailoring of the deposited ZnO composition to permit the tailoring of the ZnO film to a desired ferroelectric characteristic.
The present invention is related to methods and apparatus for processing weak ferroelectric films on semiconductor substrates. A ferroelectric film of zinc oxide (ZnO) doped with lithium (Li) and/or magnesium (Mg) is deposited on a substrate in an electron cyclotron resonance chemical vapor deposition (ECR CVD) process. An embodiment according to the invention advantageously permits the fabrication of relatively large (about 300 millimeters in diameter) substrates, which allows more devices to be fabricated at the same time, i.e., enhances economies of scale.
In accordance with one embodiment of the present invention, the zinc is introduced to a chamber through a zinc precursor in a vaporizer. Argon gas transports the zinc to the chamber. Microwave energy ionizes zinc and oxygen in the chamber to a plasma, which is directed to the substrate with a relatively strong magnetic field.
Electrically biased control grids provide a control of a rate of deposition of the plasma on the substrate. The control grids also provide Li and/or Mg dopants for the ZnO to create the ferroelectric film. The properties of the ferroelectric film can be conveniently tailored by selecting the composition, height, spacing and/or applied voltage of the control grid, which acts as a source of dopant, e.g., Li and/or Mg, for the ZnO film.